Mathematics has changed. Let's act like it

OpenAI's release of over 700 research papers on October 6 marked a significant turning point in the history of mathematics. AI has now proven capable of making groundbreaking contributions in various mathematical areas. However, some mathematicians express dissatisfaction with AI companies' apparent lack of regard for mathematical practice, citing issues like poor writing and attribution violations in the OpenAI papers.

Moreover, they argue that the effort should focus on improving how mathematicians are hired, published, and trained rather than confronting AI companies' behavior.

Some mathematicians feel threatened by the influx of AI-generated work, believing the subject they love is being "hollowed out". They assert that the announcement did not significantly impact long-standing problems like functoriality and reciprocity in the Langlands program. However, the AI's potential to uncover deeper mathematical mysteries cannot be overlooked. In the past, mathematicians have been constrained by their first natural conjectures they couldn't prove, and AI could help them venture beyond these boundaries.

The author reflects on his own journey in mathematics, noting the influence of various mathematical fields and the importance of broadening one's horizons. He hopes AI can assist mathematicians in uncovering the hidden complexities within their field. Despite the challenges posed by AI, he remains optimistic about its potential to enrich mathematics, drawing inspiration from how Apéry's proof of the irrationality of sparked his curiosity about the underlying story of algebraic functions and arithmetic.